crawling

At the moment, ItemLoader(response=response) fails if response is not a TextResponse instance.

Passing a binary response can still be useful, though. For example, to allow processors to access the response from their loader context, and hence be able to report the source URL (response.url) when reporting input issues.

Unless I missed something, the documentation doesn't explain how to query document metadata (searching "site:montferret.dev metadata" through Google returned nothing, neither did grepping the source code).

As an example, I tried to query the og:url metadata.
I tried variations of //meta[property='og:url']::attr(content), with or without the leading //, and with or without the `attr(conte

Main examples at Apify SDK webpage, Github repo and CLI templates should demonstrate how to manipulate with DOM and retrieve data from it.

Also add one example of scraping with Apify SDK + jQuery to https://sdk.apify.com/docs/examples/basiccrawler

Feedback from: https://medium.com/better-programming/do-i-need-python-scrapy-to-build-a-web-scraper-7cc7cac2081d

I lost an hour trying to make

Currently, in master, the logs are stored in a non-optional manner.

I think a good improvement for this would be to make the file logging optional, as not everyone has log rotation configured or needs logging (especially ad-hoc users).

Mentioned first here: elixir-crawly/crawly#155 (comment)

Some improvements i can think of is:

• optional switch to enabl

jsonschema latest version removed jsonschema.compat module. According to its author, this module was intended to be private and not used by external users.

We are using this module in some JSON Schema [custom validators](https://github.com/scrapinghub/spidermon/blob/master/spidermon/contrib/validation/jsonschema/form